Conventional testing systems apply loads to target computing systems based on predefined schedules. That is, a testing system may simulate activity of a certain number of users during a first predefined time period, and then, during a second predefined time period, may simulate the activity of a second number of users.